The panel featuring executives John Deere, Heritable Agriculture and crop-protection firm Invaio Sciences highlighted how AI is evolving from a set of precision tools into a software layer that connects genetics, chemistry and machines in the field. Rather than replacing farmers, panelists said, AI is absorbing complexity that growers have historically managed through experience, intuition and manual work.
From Precision Equipment to Software-Defined Farming
For decades, innovation in agriculture has centered on hardware: bigger machines, better sensors and tighter mechanical control. AI is now shifting that focus toward software-defined farming, where machines execute plans created by models that combine satellite imagery, historical yield data and real-time sensor inputs.
Melissa Neuendorf, who leads AI efforts at John Deere, pointed to harvesting as a clear example. Modern combines can now adjust their speed automatically based on predicted crop density ahead, optimizing productivity without constant human intervention.
“We’re able to predict and understand what the crop density is going to be coming up,” Neuendorf said. “Then the machine is going to adjust its speed so that it can maintain its optimum productivity.”
The goal, she emphasized, is not to turn farmers into technologists. “Farmers didn’t get into farming because they want to be data scientists,” Neuendorf said. “They got into farming because they wanted to feed and clothe the world.”
The same principle applies to weed control. PYMNTS has previously covered how John Deere has integrated AI into equipment to monitor changing field conditions.
AI Pushes Upstream Into Seeds and Chemicals
While companies like John Deere focus on execution in the field, AI is also reshaping decisions that happen years before planting. Tim Beissinger, co-founder and CTO of Heritable Agriculture, described how AI models help match plant genetics to specific environments and management practices.
“One of the biggest problems in plant genetics is genotype-by-environment interaction,” Beissinger said. “We use AI models to figure out for a particular set of genetics, where will it perform best and how will it perform best.”
Heritable is also applying newer generative AI techniques to genomes themselves, treating DNA as a language of base pairs rather than words. “This sort of thing wasn’t possible five years ago,” Beissinger said, noting that the approach accelerates discovery without necessarily requiring genetic modification.
Digital Twins, Trust and the Economics of Adoption
A recurring theme across the panel was scale. AI enables decisions to be made with a level of geographic and biological precision that was previously impractical. Beissinger said his team can now “drop a pin anywhere on the planet” and instantly estimate soil parameters and weather conditions that once required weeks of manual sampling.
That capability underpins the rise of digital twins in agriculture. John Deere’s operations platforms already allow farmers to maintain digital representations of equipment, fields and inputs. Heritable uses digital twins of plant varieties to simulate how genetics respond to changes in soil, irrigation or climate.
“We are absolutely not getting there,” Beissinger said. “We’re there.”
Still, panelists cautioned that adoption depends on trust.
“How do you build the trust of automation alongside the human experience?” Neuendorf asked. The answer, she suggested, lies in delivering clear economic value without overwhelming farmers with data they must interpret themselves.